INDEX

Explanations

mathematical expressions and programming code

The neuron is picking out tokens that are (or contain) numeric literals.

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

but

-1.60

!（

-1.27

-1.22

.],

-1.02

 with

-0.97

stubs

-0.94

by

-0.93

 before

-0.90

สามารถ

-0.90

 though

-0.89

POSITIVE LOGITS

感じが

1.05

either

1.02

0.98

事は

0.96

 either

0.95

んや

0.94

ürme

0.94

的同时

0.94

Either

0.92

 PRACTICES

0.90

Activations Density 0.166%